White Papers

24 Dell HPC Omni-Path Fabric: Supported Architecture and Application Study June 2016
5.6 CD-adapco
®
STAR-CCM+
®
At the time of publication of this whitepaper, Intel Omni-Path was not officially supported by CD-adapco
®
STAR-CCM+
®
. In order to obtain preliminary performance data for this application, the MPI software was
modified to use the appropriate Intel
®
Omni-Path library.
Multiple cases from the STAR-CCM+ benchmark suite were tested on the lab test system. The relative
performance of eight benchmark cases are presented in this section.
The graphs in Figure 16 and Figure 17 show the relative performance of the benchmarks on 1 to 32 nodes
using 28 to 896 cores. Each data point on the graphs represents the relative performance of the specific
benchmark data set using the number of cores marked on the x-axis in a parallel simulation. The results are
presented as performance relative to the performance of a single node or 28 cores. A higher value
represents better performance. The results are divided into two charts for easy readability. Figure 16 presents
benchmarks that scale relatively well up to 32 nodes and Figure 17 presents benchmarks that don’t scale as
well. For all datasets, the system scalability is as expected.
Figure 16 STAR-CCM+
®
Relative Performance Graph (1/2)
0
5
10
15
20
25
30
35
40
28 (1) 56 (2) 112 (4) 224 (8) 448 (16) 896 (32)
Performance Relative to 28 Cores (1 node)
Number of Cores (nodes)
STAR-CCM+
®
Relative Performance
civil_trim_20m EmpHydroCyclone13m HlMach10 lemans_poly_17m reactor_9m